Human loop [Flow]

# Human loop `[Flow]`

The Program maintains the model through **Research → Opinion → Merge → Position**. The human loop
orients first and escalates only when a consequential judgment is needed.

```mermaid
flowchart TD
  Run["Program run<br/>autopilot tick or run-now"] --> Orient
  Orient["Orient<br/>see what moved"] --> Consequential{Anything consequential?}
  Consequential -->|no| Caught([Caught up])
  Consequential -->|proposal awaits| Judge["Judge<br/>accept, correct, or defer"]
  Consequential -->|a question is worth reading| Steer["Steer<br/>point a read at the question"]
  Judge --> Learn
  Steer --> Learn["Learn<br/>watch a Position re-price on evidence"]
  Learn --> Refine["Refine<br/>merge a mature Stream into canon"]
  Refine --> Run
  Caught --> Run
```

Stories project the run into the human surface. They do not become a second source of truth, and model
changing actions stage a later Directive rather than claiming the work is already complete.

This is an attention-and-governance loop over the Program, not yet a complete player training loop. The
broader [Gameplay loop](gameplay-loop.md) shows how it could feed consequential practice, debrief, and
adaptation without redefining the shipped Program.

Raw Markdown source